In this paper, the real-time realization of Automatic Speech Recognition (ASR) technology on common platform is investigated. Three points are focused in the paper as following aspects. Firstly, on the base of the ASR theory, the prime principles and the methods of speech front-end detection, LPC cepstrum coefficients, acoustic modeling and its similarity estimation are studied; Secondly, some ASR programs are developed with MATLAB. Through simulations, the specificity of usual algorithms and the methods of parameter selection are compared, then a simple modeling method of column standard deviation of acoustic codebook and recognition method are presented. Lastly, based on the working above, a set of speech recognition system and the corresponding demo software are built under VC. The system is examined in real time with speaker dependent and small vocabulary. The results show that the system has a higher recognition rate. |